Team 5

Khaled Medhat Mahmoud 18P3557

Mohamed Magdy Mostafa 18P5160

import matplotlib
%matplotlib inline
import os
os.environ["KMP_DUPLICATE_LIB_OK"]="TRUE"   #This line helps with crashing
import torch
torch.cuda.empty_cache()

C:\Users\user\anaconda3\envs\env\lib\site-packages\tqdm\auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html
  from .autonotebook import tqdm as notebook_tqdm

Train a Mario-playing RL Agent

In this assignment, you’ll use RL to play fun yet challenging games. In this assignment, we’ll use RL techniques to play Mario. We’ll be using gym to make Mario environment and access it from our usual python interface as always.

You'll probably need need to install the following libraries:

1. torch
2. torchvision
3. numpy
4. gym
5. nes_py
6. gym_super_mario_bros

you can install them as instructed before.

Import important Libraries

# !pip install gym-super-mario-bros==7.3.0

import torch
from torch import nn
from torchvision import transforms as T
from PIL import Image
import numpy as np
from pathlib import Path
from collections import deque
import random, datetime, os, copy

# Gym is an OpenAI toolkit for RL
import gym
from gym.spaces import Box
from gym.wrappers import FrameStack

# NES Emulator for OpenAI Gym
from nes_py.wrappers import JoypadSpace

# Super Mario environment for OpenAI Gym
import gym_super_mario_bros

RL Definitions

---

Environment The world that an agent interacts with and learns from.

Action $a$ : How the Agent responds to the Environment. The set of all possible Actions is called action-space.

State $s$ : The current characteristic of the Environment. The set of all possible States the Environment can be in is called state-space.

Reward $r$ : Reward is the key feedback from Environment to Agent. It is what drives the Agent to learn and to change its future action. An aggregation of rewards over multiple time steps is called Return.

Optimal Action-Value function $Q^*(s,a)$ : Gives the expected return if you start in state $s$, take an arbitrary action $a$, and then for each future time step take the action that maximizes returns. $Q$ can be said to stand for the “quality” of the action in a state. We try to approximate this function.

Environment

Initialize Environment

---

In Mario, the environment consists of tubes, mushrooms and other components.

When Mario makes an action, the environment responds with the changed (next) state, reward and other info.

# Initialize Super Mario environment
env = gym_super_mario_bros.make("SuperMarioBros-1-1-v0")

# Limit the action-space to
#   0. walk right
#   1. jump right
env = JoypadSpace(env, [["right"], ["right", "A"]])

env.reset()
next_state, reward, done, info = env.step(action=0)
print(f"{next_state.shape},\n {reward},\n {done},\n {info}")

C:\Users\user\anaconda3\envs\env\lib\site-packages\gym\envs\registration.py:505: UserWarning: WARN: The environment SuperMarioBros-1-1-v0 is out of date. You should consider upgrading to version `v3` with the environment ID `SuperMarioBros-1-1-v3`.
  logger.warn(

(240, 256, 3),
 0,
 False,
 {'coins': 0, 'flag_get': False, 'life': 2, 'score': 0, 'stage': 1, 'status': 'small', 'time': 400, 'world': 1, 'x_pos': 40, 'x_pos_screen': 40, 'y_pos': 79}

Preprocess Environment

---

Environment data is returned to the agent in next_state. As you saw above, each state is represented by a [3, 240, 256] size array. Often that is more information than our agent needs; for instance, Mario’s actions do not depend on the color of the pipes or the sky!

We use Wrappers to preprocess environment data before sending it to the agent.

GrayScaleObservation is a common wrapper to transform an RGB image to grayscale; doing so reduces the size of the state representation without losing useful information. Now the size of each state: [1, 240, 256]

ResizeObservation downsamples each observation into a square image. New size: [1, 84, 84]

SkipFrame is a custom wrapper that inherits from gym.Wrapper and implements the step() function. Because consecutive frames don’t vary much, we can skip n-intermediate frames without losing much information. The n-th frame aggregates rewards accumulated over each skipped frame.

FrameStack is a wrapper that allows us to squash consecutive frames of the environment into a single observation point to feed to our learning model. This way, we can identify if Mario was landing or jumping based on the direction of his movement in the previous several frames.

class SkipFrame(gym.Wrapper):
    def __init__(self, env, skip):
        """Return only every `skip`-th frame"""
        super().__init__(env)
        self._skip = skip

    def step(self, action):
        """Repeat action, and sum reward"""
        total_reward = 0.0
        done = False
        for i in range(self._skip):
            # Accumulate reward and repeat the same action
            obs, reward, done, info = self.env.step(action)
            total_reward += reward
            if done:
                break
        return obs, total_reward, done, info


class GrayScaleObservation(gym.ObservationWrapper):
    def __init__(self, env):
        super().__init__(env)
        obs_shape = self.observation_space.shape[:2]
        self.observation_space = Box(low=0, high=255, shape=obs_shape, dtype=np.uint8)

    def permute_orientation(self, observation):
        # permute [H, W, C] array to [C, H, W] tensor
        observation = np.transpose(observation, (2, 0, 1))
        observation = torch.tensor(observation.copy(), dtype=torch.float)
        return observation

    def observation(self, observation):
        observation = self.permute_orientation(observation)
        transform = T.Grayscale()
        observation = transform(observation)
        return observation


class ResizeObservation(gym.ObservationWrapper):
    def __init__(self, env, shape):
        super().__init__(env)
        if isinstance(shape, int):
            self.shape = (shape, shape)
        else:
            self.shape = tuple(shape)

        obs_shape = self.shape + self.observation_space.shape[2:]
        self.observation_space = Box(low=0, high=255, shape=obs_shape, dtype=np.uint8)

    def observation(self, observation):
        transforms = T.Compose(
            [T.Resize(self.shape), T.Normalize(0, 255)]
        )
        observation = transforms(observation).squeeze(0)
        return observation


# Apply Wrappers to environment
env = SkipFrame(env, skip=4)
env = GrayScaleObservation(env)
env = ResizeObservation(env, shape=84)
env = FrameStack(env, num_stack=4)

After applying the above wrappers to the environment, the final wrapped state consists of 4 gray-scaled consecutive frames stacked together, as shown above in the image on the left. Each time Mario makes an action, the environment responds with a state of this structure. The structure is represented by a 3-D array of size [4, 84, 84].

Agent

---

We create a class Mario to represent our agent in the game. Mario should be able to:

• Act according to the optimal action policy based on the current state (of the environment).

• Remember experiences. Experience = (current state, current action, reward, next state). Mario caches and later recalls his experiences to update his action policy.

• Learn a better action policy over time

class Mario:
    def __init__():
        pass

    def act(self, state):
        """Given a state, choose an epsilon-greedy action"""
        #TODO: """You'll need to implement this function. see below for actual implementation"""
        pass

    def cache(self, experience):
        """Add the experience to memory"""
        #TODO: """You'll need to implement this function. see below for actual implementation"""
        pass

    def recall(self):
        """Sample experiences from memory"""
        #TODO: """You'll need to implement this function. see below for actual implementation"""
        pass

    def learn(self):
        """Update online action value (Q) function with a batch of experiences"""
        #TODO: """You'll need to implement this function. see below for actual implementation"""
        pass

In the following sections, we will populate Mario’s parameters and define his functions.

Act

---

For any given state, an agent can choose to do the most optimal action (exploit) or a random action (explore).

Mario randomly explores with a chance of self.exploration_rate; when he chooses to exploit, he relies on MarioNet (implemented in Learn section) to provide the most optimal action.

class Mario:
    def __init__(self, state_dim, action_dim, save_dir):
        self.state_dim = state_dim
        self.action_dim = action_dim
        self.save_dir = save_dir

        self.use_cuda = torch.cuda.is_available()

        # Mario's DNN to predict the most optimal action - we implement this in the Learn section
        self.net = MarioNet(self.state_dim, self.action_dim).float()
        if self.use_cuda:
            self.net = self.net.to(device="cuda")

        self.exploration_rate = 0.8
        self.exploration_rate_decay = 0.99999975
        self.exploration_rate_min = 0.1
        self.curr_step = 0

        self.save_every = 100000  # no. of experiences between saving Mario Net

    def act(self, state):
        """
    Given a state, choose an epsilon-greedy action and update value of step.

    Inputs:
    state(LazyFrame): A single observation of the current state, dimension is (state_dim)
    Outputs:
    action_idx (int): An integer representing which action Mario will perform
    """
        # EXPLORE
        if np.random.rand() < self.exploration_rate:
            action_idx = np.random.randint(self.action_dim)

        # EXPLOIT
        else:
            state = state.__array__()
            if self.use_cuda:
                #TODO1: convert the state numpy array to torch tensor. Then, move that tensor to cuda
                state = torch.FloatTensor(state).cuda() 
            else:
                #TODO2: convert the state numpy array to torch tensor. DO NOT move that tensor to cuda
                state = torch.FloatTensor(state)
            
            #TODO3: use the neural network to get the next action from the given state. 
            state = state.unsqueeze(0)                                #Returns a tensor of size 1 
            action_values = self.net(state, model="online")           #Get action values     
            action_idx = torch.argmax(action_values, axis=1).item()   #Get best action index from action values


        #TODO4: decrease exploration_rate
        self.exploration_rate *= self.exploration_rate_decay
        self.exploration_rate = max(self.exploration_rate_min, self.exploration_rate)
        # increment step
        self.curr_step += 1
        return action_idx

Cache and Recall

These two functions serve as Mario’s “memory” process.

cache(): Each time Mario performs an action, he stores the experience to his memory. His experience includes the current state, action performed, reward from the action, the next state, and whether the game is done.

recall(): Mario randomly samples a batch of experiences from his memory, and uses that to learn the game.

class Mario(Mario):  # subclassing for continuity
    def __init__(self, state_dim, action_dim, save_dir):
        super().__init__(state_dim, action_dim, save_dir)
        self.memory = deque(maxlen=100000)
        self.batch_size = 32

    def cache(self, state, next_state, action, reward, done):
        """
        Store the experience to self.memory (replay buffer)

        Inputs:
        state (LazyFrame),
        next_state (LazyFrame),
        action (int),
        reward (float),
        done(bool))
        """
        state = state.__array__()
        next_state = next_state.__array__()

        if self.use_cuda:
            #TODO5: Convert the state, next_state, action, reward, and done to torch tensor. Move them to cuda to be 
            #      stored in memory later 
            state = torch.FloatTensor(state).cuda()
            next_state = torch.FloatTensor(next_state).cuda()
            action = torch.LongTensor([action]).cuda() 
            reward = torch.FloatTensor([reward]).cuda() 
            done = torch.BoolTensor([done]).cuda()
        else:
            #TODO6: Convert the state, next_state, action, reward, and done to torch tensor. Do NOT to cuda to be 
            #      stored in memory later 
            state = torch.FloatTensor(state)
            next_state = torch.FloatTensor(next_state)
            action = torch.LongTensor([action])
            reward = torch.FloatTensor([reward])
            done = torch.BoolTensor([done])

        self.memory.append((state, next_state, action, reward, done,))

    def recall(self):
        """
        Retrieve a batch of experiences from memory
        """
        #TODO: Sample state, next_state, action, reward, and done from our memory (replay buffer)
        
        batch = random.sample(self.memory, self.batch_size)
        state, next_state, action, reward, done = map(torch.stack, zip(*batch))
        return state, next_state, action.squeeze(), reward.squeeze(), done.squeeze()

Learn

---

Mario uses the DDQN algorithm under the hood. DDQN uses two ConvNets - $Q_{online}$ and $Q_{target}$ - that independently approximate the optimal action-value function.

In our implementation, we share feature generator features across $Q_{online}$ and $Q_{target}$, but maintain separate FC classifiers for each. $\theta_{target}$ (the parameters of $Q_{target}$) is frozen to prevent updation by backprop. Instead, it is periodically synced with $\theta_{online}$ (more on this later).

class MarioNet(nn.Module):
    """mini cnn structure
  input -> (conv2d + relu) x 3 -> flatten -> (dense + relu) x 2 -> output
  """

    def __init__(self, input_dim, output_dim):
        super().__init__()
        c, h, w = input_dim

        if h != 84:
            raise ValueError(f"Expecting input height: 84, got: {h}")
        if w != 84:
            raise ValueError(f"Expecting input width: 84, got: {w}")

        self.online = nn.Sequential(
            nn.Conv2d(in_channels=c, out_channels=32, kernel_size=8, stride=4),
            nn.ReLU(),
            nn.Conv2d(in_channels=32, out_channels=64, kernel_size=4, stride=2),
            nn.ReLU(),
            nn.Conv2d(in_channels=64, out_channels=64, kernel_size=3, stride=1),
            nn.ReLU(),
            nn.Flatten(),
            nn.Linear(3136, 512),
            nn.ReLU(),
            nn.Linear(512, output_dim),
        )

        self.target = copy.deepcopy(self.online)

        # Q_target parameters are frozen.
        for p in self.target.parameters():
            p.requires_grad = False

    def forward(self, input, model):
        if model == "online":
            return self.online(input)
        elif model == "target":
            return self.target(input)

TD Estimate & TD Target

---

Two values are involved in learning:

TD Estimate - the predicted optimal $Q^*$ for a given state $s$

\begin{align}{TD}_e = Q_{online}^*(s,a)\end{align}

TD Target - aggregation of current reward and the estimated $Q^*$ in the next state $s'$

\begin{align}a' = argmax_{a} Q_{online}(s', a)\end{align}\begin{align}{TD}_t = r + \gamma Q_{target}^*(s',a')\end{align}

Because we don’t know what next action $a'$ will be, we use the action $a'$ maximizes $Q_{online}$ in the next state $s'$.

Notice we use the @torch.no_grad() decorator on td_target() to disable gradient calculations here (because we don’t need to backpropagate on $\theta_{target}$).

class Mario(Mario):
    def __init__(self, state_dim, action_dim, save_dir):
        super().__init__(state_dim, action_dim, save_dir)
        self.gamma = 0.9

    def td_estimate(self, state, action):
        current_Q = self.net(state, model="online")[
            np.arange(0, self.batch_size), action
        ]  # Q_online(s,a)
        return current_Q

    @torch.no_grad()
    def td_target(self, reward, next_state, done):
        next_state_Q = self.net(next_state, model="online")
        best_action = torch.argmax(next_state_Q, axis=1)
        next_Q = self.net(next_state, model="target")[
            np.arange(0, self.batch_size), best_action
        ]
        return (reward + (1 - done.float()) * self.gamma * next_Q).float()

Updating the model

---

As Mario samples inputs from his replay buffer, we compute $TD_t$ and $TD_e$ and backpropagate this loss down $Q_{online}$ to update its parameters $\theta_{online}$ ($\alpha$ is the learning rate lr passed to the optimizer)

\begin{align}\theta_{online} \leftarrow \theta_{online} + \alpha \nabla(TD_e - TD_t)\end{align}

$\theta_{target}$ does not update through backpropagation. Instead, we periodically copy $\theta_{online}$ to $\theta_{target}$

\begin{align}\theta_{target} \leftarrow \theta_{online}\end{align}

class Mario(Mario):
    def __init__(self, state_dim, action_dim, save_dir):
        super().__init__(state_dim, action_dim, save_dir)
        self.optimizer = torch.optim.Adam(self.net.parameters(), lr=0.00025)
        self.loss_fn = torch.nn.SmoothL1Loss()

    def update_Q_online(self, td_estimate, td_target):
        loss = self.loss_fn(td_estimate, td_target)
        self.optimizer.zero_grad()
        loss.backward()
        self.optimizer.step()
        return loss.item()

    def sync_Q_target(self):
        self.net.target.load_state_dict(self.net.online.state_dict())

Save checkpoint ~~~~~~

class Mario(Mario):
    def save(self):
        save_path = (
            self.save_dir / f"mario_net_{int(self.curr_step // self.save_every)}.chkpt"
        )
        torch.save(
            dict(model=self.net.state_dict(), exploration_rate=self.exploration_rate),
            save_path,
        )
        print(f"MarioNet saved to {save_path} at step {self.curr_step}")

Putting it all together ~~~~~~

class Mario(Mario):
    def __init__(self, state_dim, action_dim, save_dir):
        super().__init__(state_dim, action_dim, save_dir)
        self.burnin = 1e4  # min. experiences before training
        self.learn_every = 3  # no. of experiences between updates to Q_online
        self.sync_every = 1e4  # no. of experiences between Q_target & Q_online sync

    def learn(self):
        if self.curr_step % self.sync_every == 0:
            self.sync_Q_target()

        if self.curr_step % self.save_every == 0:
            self.save()

        if self.curr_step < self.burnin:
            return None, None

        if self.curr_step % self.learn_every != 0:
            return None, None

        #TODO6: Sample from memory
        state, next_state, action, reward, done = self.recall() 

        # Get TD Estimate
        td_est = self.td_estimate(state, action)

        # Get TD Target
        td_tgt = self.td_target(reward, next_state, done)

        # Backpropagate loss through Q_online
        loss = self.update_Q_online(td_est, td_tgt)

        return (td_est.mean().item(), loss)

Logging

import numpy as np
import time, datetime
import matplotlib.pyplot as plt


class MetricLogger:
    def __init__(self, save_dir):
        self.save_log = save_dir / "log"
        with open(self.save_log, "w") as f:
            f.write(
                f"{'Episode':>8}{'Step':>8}{'Epsilon':>10}{'MeanReward':>15}"
                f"{'MeanLength':>15}{'MeanLoss':>15}{'MeanQValue':>15}"
                f"{'TimeDelta':>15}{'Time':>20}\n"
            )
        self.ep_rewards_plot = save_dir / "reward_plot.jpg"
        self.ep_lengths_plot = save_dir / "length_plot.jpg"
        self.ep_avg_losses_plot = save_dir / "loss_plot.jpg"
        self.ep_avg_qs_plot = save_dir / "q_plot.jpg"

        # History metrics
        self.ep_rewards = []
        self.ep_lengths = []
        self.ep_avg_losses = []
        self.ep_avg_qs = []

        # Moving averages, added for every call to record()
        self.moving_avg_ep_rewards = []
        self.moving_avg_ep_lengths = []
        self.moving_avg_ep_avg_losses = []
        self.moving_avg_ep_avg_qs = []

        # Current episode metric
        self.init_episode()

        # Timing
        self.record_time = time.time()

    def log_step(self, reward, loss, q):
        self.curr_ep_reward += reward
        self.curr_ep_length += 1
        if loss:
            self.curr_ep_loss += loss
            self.curr_ep_q += q
            self.curr_ep_loss_length += 1

    def log_episode(self):
        "Mark end of episode"
        self.ep_rewards.append(self.curr_ep_reward)
        self.ep_lengths.append(self.curr_ep_length)
        if self.curr_ep_loss_length == 0:
            ep_avg_loss = 0
            ep_avg_q = 0
        else:
            ep_avg_loss = np.round(self.curr_ep_loss / self.curr_ep_loss_length, 5)
            ep_avg_q = np.round(self.curr_ep_q / self.curr_ep_loss_length, 5)
        self.ep_avg_losses.append(ep_avg_loss)
        self.ep_avg_qs.append(ep_avg_q)

        self.init_episode()

    def init_episode(self):
        self.curr_ep_reward = 0.0
        self.curr_ep_length = 0
        self.curr_ep_loss = 0.0
        self.curr_ep_q = 0.0
        self.curr_ep_loss_length = 0

    def record(self, episode, epsilon, step):
        mean_ep_reward = np.round(np.mean(self.ep_rewards[-100:]), 3)
        mean_ep_length = np.round(np.mean(self.ep_lengths[-100:]), 3)
        mean_ep_loss = np.round(np.mean(self.ep_avg_losses[-100:]), 3)
        mean_ep_q = np.round(np.mean(self.ep_avg_qs[-100:]), 3)
        self.moving_avg_ep_rewards.append(mean_ep_reward)
        self.moving_avg_ep_lengths.append(mean_ep_length)
        self.moving_avg_ep_avg_losses.append(mean_ep_loss)
        self.moving_avg_ep_avg_qs.append(mean_ep_q)

        last_record_time = self.record_time
        self.record_time = time.time()
        time_since_last_record = np.round(self.record_time - last_record_time, 3)

        print(
            f"Episode {episode} - "
            f"Step {step} - "
            f"Epsilon {epsilon} - "
            f"Mean Reward {mean_ep_reward} - "
            f"Mean Length {mean_ep_length} - "
            f"Mean Loss {mean_ep_loss} - "
            f"Mean Q Value {mean_ep_q} - "
            f"Time Delta {time_since_last_record} - "
            f"Time {datetime.datetime.now().strftime('%Y-%m-%dT%H:%M:%S')}"
        )

        with open(self.save_log, "a") as f:
            f.write(
                f"{episode:8d}{step:8d}{epsilon:10.3f}"
                f"{mean_ep_reward:15.3f}{mean_ep_length:15.3f}{mean_ep_loss:15.3f}{mean_ep_q:15.3f}"
                f"{time_since_last_record:15.3f}"
                f"{datetime.datetime.now().strftime('%Y-%m-%dT%H:%M:%S'):>20}\n"
            )

        for metric in ["ep_rewards", "ep_lengths", "ep_avg_losses", "ep_avg_qs"]:
            plt.plot(getattr(self, f"moving_avg_{metric}"))
            plt.savefig(getattr(self, f"{metric}_plot"))
            plt.clf()

Let’s play!

---

In this example we run the training loop for 10 episodes, but for Mario to truly learn the ways of his world, we suggest running the loop for at least 40,000 episodes!

use_cuda = torch.cuda.is_available()
print(f"Using CUDA: {use_cuda}")
print()

save_dir = Path("checkpoints") / datetime.datetime.now().strftime("%Y-%m-%dT%H-%M-%S")
save_dir.mkdir(parents=True)

mario = Mario(state_dim=(4, 84, 84), action_dim=env.action_space.n, save_dir=save_dir)

logger = MetricLogger(save_dir)

episodes = 21

Using CUDA: True

Changed the location of env.close() in order to be able to watch the agent through each episode

np.seterr(over='ignore')
PATH = 'F:\Mario Deep Learning Project\saving\MarioModel'
mario.net.load_state_dict(torch.load(PATH))

for e in range(episodes):
    state=env.reset()
    # Play the game!
    while True:
        env.render()
        # Run agent on the state
        action = mario.act(state)
        # Agent performs action
        next_state, reward, done, info = env.step(action)

        # Remember
        mario.cache(state, next_state, action, reward, done)

        # Learn
        q, loss = mario.learn()

        # Logging
        logger.log_step(reward, loss, q)
        # Update state
        state = next_state
        
        # Check if end of game
        if done or info["flag_get"]:
            break
    
    logger.log_episode()
    if e % 5 == 0:
        logger.record(episode=e, epsilon=mario.exploration_rate, step=mario.curr_step)
        
    if (e + 1)%20 == 0:
          torch.save(mario.net.state_dict(), PATH)
        

<>:2: DeprecationWarning: invalid escape sequence \M
C:\Users\user\anaconda3\envs\env\lib\site-packages\pyglet\image\codecs\wic.py:289: UserWarning: [WinError -2147417850] Cannot change thread mode after it is set
  warnings.warn(str(err))
C:\Users\user\AppData\Local\Temp\ipykernel_62760\363398287.py:28: DeprecationWarning: In future, it will be an error for 'np.bool_' scalars to be interpreted as an index
  done = torch.BoolTensor([done]).cuda()

Episode 0 - Step 157 - Epsilon 0.7999686006122877 - Mean Reward 636.0 - Mean Length 157.0 - Mean Loss 0.0 - Mean Q Value 0.0 - Time Delta 5.091 - Time 2022-05-31T18:50:25
Episode 5 - Step 1792 - Epsilon 0.7996416802247831 - Mean Reward 893.5 - Mean Length 298.667 - Mean Loss 0.0 - Mean Q Value 0.0 - Time Delta 17.971 - Time 2022-05-31T18:50:43
Episode 10 - Step 2672 - Epsilon 0.7994657783830337 - Mean Reward 772.364 - Mean Length 242.909 - Mean Loss 0.0 - Mean Q Value 0.0 - Time Delta 10.007 - Time 2022-05-31T18:50:53
Episode 15 - Step 4345 - Epsilon 0.7991314716961484 - Mean Reward 794.062 - Mean Length 271.562 - Mean Loss 0.0 - Mean Q Value 0.0 - Time Delta 18.665 - Time 2022-05-31T18:51:12
Episode 20 - Step 4980 - Epsilon 0.7990046196282916 - Mean Reward 719.857 - Mean Length 237.143 - Mean Loss 0.0 - Mean Q Value 0.0 - Time Delta 7.391 - Time 2022-05-31T18:51:19

<Figure size 432x288 with 0 Axes>

model_save_name = 'MarioModel'
path = f"F:\Mario Deep Learning Project\saving\\{model_save_name}"

mario = Mario(state_dim=(4, 84, 84), action_dim=env.action_space.n, save_dir=save_dir)

mario.net.load_state_dict(torch.load(path))

# Limit the action-space to
#   0. walk right
#   1. jump right
for i in range(20):
    state = env.reset()
    while True:
        # Run agent on the state
        env.render()
        action = mario.act(state)
        # Agent performs action
        next_state, reward, done, info = env.step(action)

        # Update state
        state = next_state
        
        # Check if end of game
        if done or info["flag_get"]:
            break
env.close()

Conclusion

---

In this tutorial, we saw how we can use PyTorch to train a game-playing AI. You can use the same methods to train an AI to play any of the games at the OpenAI gym <https://gym.openai.com/>